Choosing brands: fresh produce versus other products.

Brands tend to generate significant premiums, and thus their uses have been considered in enhancing the value added of farming (Hayes and Lence 2002). Yet, brands are less commonly used in fresh agricultural produce than other products (Kaufman et al. 2000). With the greater emphasis on product differentiation, research on agricultural marketing strategies requires understanding the reasons for the relative paltry use of brands in the farm sector. In addition, it is important to assess the relative gains from introducing brands of fresh agricultural products and to identify features of their likely buyers. This article develops a methodological and empirical strategy to answer these questions and applies it to data collected in College Station and Bryan, Texas, in fall 2006.

We will investigate the following questions to explain the lack of brands of fresh produce: (a) Do the same causes of preferring brands over generics apply to fresh produce? (b) Do consumers have significantly lower willingness to pay (WTP) for brands of fresh produce than other product categories? (c) To what extent are brands' premiums and market shares of produce different from other products? (d) To what extent are consumers consistent in their brand preference across product categories?

The rest of this article is organized as follows. We present a simple framework to explain different components that contribute to the brand premium. We then provide data information and discuss the empirical results. Our estimation results show that (a) consumers have a lower WTP for brands of fresh produce than in other categories including electronics, clothing, and packaged food; and (b) certain socio-demographic factors play an important role in WTP for brands, including income, education, age, race, gender, and household size. The simulation results suggest that brands of fresh produce have a higher optimal price premium but a much smaller market share than those of other products. The simulation results also show that individuals are consistent in their WTP for brands across product categories, and thus there is a potential gain from selling brands of fresh fruits and vegetables in outlets selling brands of other products.

The Simple Brand Value Equation and the Basic Hypothesis

Adopting Rosen's hedonic pricing methodology (Rosen 1974), we introduce a simple formulation to compare the relative gain of brand products over generic across different product categories. The term, "the value of the product to a consumer," is used here to denote the consumer's WTP for the product, i.e., the price that will make the consumer indifferent to purchasing or not purchasing the product. The value of the product to a consumer is a function of the features of the product and the characteristics of the buyer. The literature suggests that branded products are more valuable because consumers associate them with better performance in three key areas:

* Quality/reliability. Erdem, Zhao, and Valenzuela (2004) find that both the perceived average and variability levels of quality explain the premium received by national brands over store brands for different products across countries.

* Design. Brands may have more attractive appearance and better performance related to design than generic products. Vranesevic and Stancec (2003) suggest that appearance is viewed as a distinct characteristic of a food brand.

* Prestige. Aaker and Joachimsthaler (2002) find that brands provide "self-expressive benefits," as the association with brands contributes to the buyer's self-image.

Let [DELTA]V be the value difference to consumers between a brand product and a generic product, i.e., the perceived extra value of the brand product relative to the generic product's original value, V. We assume that this difference can be simply decomposed into three hedonic components relating to quality/reliability, design, and prestige:

(1) [DELTA]V = [DELTA]D + [DELTA]P + [DELTA]Q.

[DELTA]D is the extra value attributed to improved design of the brand product, i.e., a better design contributes to more attractive appearance and/or better functionality of the product. [DELTA]P is the added value reflecting the prestige added by a brand. The value of the extra quality added by the product is denoted by [DELTA]Q. We narrowly interpret quality to mean improved reliability, reflecting risk reduction due to a lower probability of product failure and a lower loss in case of failure. The quality gain can be further decomposed to

(2) [DELTA]Q = [q.sub.G][L.sub.G] - [q.sub.B][L.sub.B] = [DELTA]q[L.sub.B] + [DELTA]l[q.sub.G]

where [q.sub.G] and [q.sub.B] are the product failure probabilities, and [L.sub.G] and [L.sub.B] are the losses after product failure for the generic and the brand varieties, respectively. It is plausible to assume that brands reduce the probability of loss, and this reduction is [DELTA]q = [q.sub.G] - [q.sub.B] > 0. Brands are likely to cause less loss in case of product failure since they tend to have a better product support, for example, better warranties or money-back guarantee programs, thus we assume [DELTA]L = [L.sub.G] - [L.sub.B] > 0. Therefore, brands provide better quality in the sense of a reliable product, so [DELTA]Q > 0. Similarly, we assume brands provide extra value in terms of prestige and design. However, the relative gains from brands vary across product categories. We rely on prior studies to hypothesize on the relative value gains from brands for the four product categories we consider, in terms of value in quality, design, and prestige effects.

The relative contributions of the brand's quality effect. Two factors determine the quality gain from buying brands of products, ex ante consumer learning and durability. Ex ante consumer learning, where consumers use demonstrations and in-store tests to reduce product quality uncertainty, is a partial substitute to brands in providing information about product quality. Brands have significant value in providing information about quality of experience goods when ex ante learning is limited, as is the case with electronics and packaged foods, in contrast to fresh produce and clothing. The second factor is durability. Caves and Greene (1996) state that the value of brands as a quality signal is larger for durables because buyers cannot frequently adjust purchasing behavior. Electronics and clothing are durable experience products, while food products are frequently purchased items; hence, the brand effect for electronics and clothing is greater than for food. Therefore, brands provide extra value as quality signals for electronics both because they are durable goods and because of limited ex ante consumer learning. For clothing, brands convey quality information mostly because of durability, but provide limited ex ante consumer learning because clothing is tried on pre-purchase. In the case of packaged food products, the quality value of brands stems from the lack of effective ex ante means to assess quality, but is limited by the non-durable nature of the product. Both lack of durability and high availability of ex ante learning reduce the value of brands in fresh produce. Based on this analysis, we conjecture

(3) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

The relative contributions of the brand's design effect. Design and appearance are major attributes of brand fashion products (Moore, Fernie, and Burt 2000). The additional value gain in brand electronics is stronger when products are differentiated in their external design (Holbrook 1992). Fresh fruits and vegetables can be "designed" by plant breeding and cultural practices, and design features like size and color strongly affect produce prices (Parker and Zilberman 1993), but we could not find evidence that consumers associate better design with produce brands. The above suggests

(4) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

The relative contributions of the brand's prestige effect. Auty and Elliott (1999) find that brand fashion products affect the self-image of buyers, and Holbrook (1992) argues that image effects contribute to the value of brands in electronics to the extent that consumption is seen by others. Thus, clothing will likely provide the most prestige, as it has the most exposure to other people, followed by electronic gadgets. Brand products of food items are the least valuable source of prestige. We are aware that consumers may convey certain images by buying organic food, fair-trade food, etc., but the prestige impact is certainly much lower than for clothing or electronics. This discussion suggests an inequality similar to equation (4), namely,

(5) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

The results of inequalities relating to three dimensions of brand value to consumers allow some comparison of the relative contributions of brands to the product values. They suggest that brands make relatively the least contribution for fresh produce, and for packaged foods the contribution of brands to value is likely to be smaller than in electronics and clothing, i.e.,

(6) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].